Issue with loopback interface with remote mirroring

Hi all.
I configured remote mirroring according to "EXOS_User_Guide_16_1".
Remote mirror seems to work but i got BPDU error on intereface assigned as loopback for mirroring.
mirroring config:

create mirror "voip-span"
configure mirror voip-span to port-list 51,52 loopback-port 48 remote-tag 1000
configure mirror voip-span add vlan "PHONE" ingress
enable mirror voip-span

error message:

02/25/2016 11:21:24.02 : Port=48: Recv own loop back packet from same port
02/25/2016 11:21:26.01 : Port=48: Recv own loop back packet from same port
02/25/2016 11:21:28.02 : Port=48: Recv own loop back packet from same port

i got this on every switch i use as mirror source.

System Type: X460G2-48p-10G4
Image: ExtremeXOS version by release-manager

During that error i got disruption with connectivity - every few pings are lost.
How can i resolve that ? Any ideas ?

7 replies

Userlevel 6
Hi Marek,

Is there any configuration (Vlan, STP, etc) in the loopback port 48?
No nothing is configured. I believe STP is added by mirroring configuration.

# show vlan ports 48
Error: The specified ports are not members of any VLANs.

# show "s0" ports 48
Port Mode State Cost Flags Priority Port ID Designated Bridge
48 802.1D FORWARDING 20000 eDap-w---- 128 8030 80:00:00:04:96:99:83:e8

Total Ports: 1

------------------------- Flags: ----------------------------
1: e=Enable, d=Disable
2: (Port role) R=Root, D=Designated, A=Alternate, B=Backup, M=Master
3: (Config type) b=broadcast, p=point-to-point, e=edge, a=auto
4: (Oper. type) b=broadcast, p=point-to-point, e=edge
5: p=proposing, a=agree
6: (partner mode) d = 802.1d, w = 802.1w, m = mstp
7: i = edgeport inconsistency
8: S = edgeport safe guard active
s = edgeport safe guard configured but inactive
8: G = edgeport safe guard bpdu restrict active in 802.1w and mstp
g = edgeport safe guard bpdu restrict active in 802.1d
9: B = Boundary, I = Internal
10: r = restricted role, t = active role

that is avlan that created itself when remote mirror was added

# show vlan "internalMirrorLoopback_0"
VMAN Interface for mirroring with name internalMirrorLoopback_0 created by user
Admin State: Enabled Tagging:Untagged (Internal tag 4092)
Description: None
Virtual router: VR-Default
IPv4 Forwarding: Disabled
IPv4 MC Forwarding: Disabled
IPv6 Forwarding: Disabled
IPv6 MC Forwarding: Disabled
IPv6: None
STPD: s0(Enabled,Auto-bind)
Protocol: Match all unfiltered protocols
Loopback: Disabled
NetLogin: Disabled
OpenFlow: Disabled
TRILL: Disabled
QosProfile: None configured
Egress Rate Limit Designated Port: None configured
Flood Rate Limit QosProfile: None configured
Ports: 3. (Number of active ports=3)
Untag: *48L, *51, *52
Flags: (*) Active, (!) Disabled, (g) Load Sharing port
(b) Port blocked on the vlan, (m) Mac-Based port
(a) Egress traffic allowed for NetLogin
(u) Egress traffic unallowed for NetLogin
(t) Translate VLAN tag for Private-VLAN
(s) Private-VLAN System Port, (L) Loopback port
(x) VMAN Tag Translated port
(G) Multi-switch LAG Group port
(H) Dynamically added by MVRP
(D) TRILL Designated, (A) TRILL Appointed Forwarder
(I) Dynamically added by IDM
(U) Dynamically added uplink port
(V) Dynamically added by VM Tracking

Port 51,52 are uplink port to other switches.

Userlevel 6
Hi Marek, are you using STP for remote-mirroring to have redundancy as described in the EXOS UserGuide? Can you please share the topology?

Try to remove the loopback port from S0 STPD using the command below and please share the results:

configure s0 delete vlan internalMirrorLoopback_0 ports 48
Topology in short looks like this with "access switch loop to core" multiplicated on 3 different floors:

When I remove loopback from stp errors stop showing up.
But now i have other issue - when LAG or even single link in lag between access switches is connected i got 50-60% cpu usage on all switches in network ! even on those on server side - so it looks like some storm or something. in TOP I just see bcmRX to use as mush as 40-50% of cpu.
And it is like this only on one floor.
I'm looking into STP but cannot find an issue. I think i'll need to first figure out what is the issue with that connection and then take care of that mirroring.
Userlevel 6
Hi Marek,

That behavior (CPU utilization / bcmRX) indicates looping and it needs to be fixed prior to the remote-mirroring configuration.

Based on this topology, if all those devices are EXOS platform I would recommend you to enable EAPS instead of STP.

How many vlans have you extended among those 6 devices? Also, can you please confirm if all those units are EXOS platforms?
Yes all are exos devices.
I'll go through all STP and vlan config once again on all switches and than we'll see if problem with mirroring is still occuring.

Thanks Henrique.
Userlevel 7
Marek T wrote:

Yes all are exos devices.
I'll go through all STP and vlan config once again on all switches and than we'll see if problem with mirroring is still occuring.

Thanks Henrique.

Hi Marek, were you able to get this resolved?